telvis07 / twitter_mining

The goal of this project is to develop novel ways find the MOST meaningful tweets and tweeters over a given interval.

Twitter Mining

This project contains python code for twitter mining.

Goal

My goal is to find novel ways to summarize topics and trends on twitter.

Dependencies

• Python
• CouchDB
• tweepy

To run

1. Run filter_tweets_streaming_api.py to load tweets to couchdb. The delay shows when your consumer is too slow.

   $ python filter_tweets_streaming_api.py test twitter
   opened test
   Track parameters ['twitter']
   Delay 0 seconds. created_at 2012-03-05 01:11:45. id_str '17647.....'. tweeter 'sometweeter'. tweet 'I love to tweet!'
   Delay 0 seconds. created_at 2012-03-05 01:11:45. id_str '17647.....'. tweeter 'tweeter_guy'. tweet 'Twitter is fun!'
   ...

2. Run views.py to create indexes on your couchdb database
   $ python views.py test

3. Run a "reporting" script such as top_tweeters_by_follower_count.py and send a summary email of tweets
   $ python top_tweeters_by_follower_count.py -d test 2012-03-05 --dry-run
   Top 10 tweeters
   tweeter1
   "look at my awesome tweet"
   tweeter2
   "i'm colder than a polar bears toe nails"

4. That's all for now...